AI courses

Browse public AI courses with structured lessons, visuals, and practice on Wondering.

• World Models: How AI Simulates Worlds

  For curious learners who want to understand AI world models: what they are, how systems like Genie, V-JEPA, and Dreamer learn to predict what happens next, where the idea came from, who is building them now, and where they still break.

• 产品经理的AI实战课

  专为产品经理打造，带你用通俗且硬核的视角搞懂底层逻辑，把前沿AI与Agent技术直接落地到日常规划、原型设计与功能交付中。 * **AI 原理与边界洞察**：搞懂底层逻辑，精准定义可行需求 * **AI 原生产品设计**：掌握 Prompt 工程与上下文构建技巧 * **智能体实战构建**：独立设计并落地自动化 Agent 流程 * **AI 驱动敏捷交付**：运用工具链极速完成原型设计与发版

• iPAS 中級考前衝刺（數據分析）

  聚焦統計與 Pandas 核心關鍵字，助你快速突破弱點，雙科穩拿 70 分。 * **機率統計與推論**：速解核心考點 * **Pandas 程式判讀**：精準抓出語法關鍵字 * **中級術語直覺反射**：秒殺二選一干擾選項 * **L21 與 L22 雙科**：穩拿 70 分過標

• Jev System One Essentials

  Master the one-model architecture by exploring its components, data flow, and trade-offs. You will gain the practical skills needed to confidently build, configure, and debug systems. * **Master one-model architecture**: core components, concepts, and data flow * **Evaluate design trade-offs**: make informed architectural decisions for performance * **Build and configure systems**: hands-on implementation of developer workflows * **Debug production issues**: troubleshoot data flow and system integration errors

• Why Your RAG System Gives Bad Answers

  For builders and technical product managers diagnosing document-grounded answers, a single policy question exposes failures in chunking, retrieval, reranking, context construction and answering. Distinguish missing evidence from ignored evidence, then choose targeted fixes and checks.

• Delegating Work to AI

  For managers and knowledge workers who already use AI, this course turns repeatable tasks into clear briefs with relevant context, acceptance criteria, and review checkpoints. It shows how to check outputs, handle missing information, and keep consequential actions under named human approval.

• Claude Code for Product Work

  For product managers and operators comfortable with AI chat, this course connects synthetic feedback analysis, evidence-backed specifications, and a small local prototype. You will check behavior against requirements, inspect file changes, and manage permissions and data deliberately.

• AI Product Decisions: From Idea to Launch

  For product managers and founders assessing AI features. Choose a suitable user problem, write a bounded feature specification with acceptance criteria, justify prompting, retrieval or fine-tuning, and balance quality, speed, cost and launch risk.

• Building Reliable AI Agents

  For builders and technical product managers who know prompting, a damaged-item refund shows when to use a fixed workflow or an agent. Tools, context, permissions, bounded recovery, execution records, and human handoffs make each next action easier to trust.

• AI Evals: Measuring Product Quality

  For product managers and engineers who already use LLMs, this course connects real output failures to representative tests, task-specific rubrics, and human-calibrated judges. Compare product changes with visible regressions and clear limits on what the scores prove.

• Stanford CS336 Language Modeling from Scratch | Spring 2026 | Lecture 1: Overview, Tokenization

  Percy Liang's opening Stanford CS336 lecture explains why the course builds language models from scratch under a compute budget, then works through tokenizers: characters, bytes, words, and byte pair encoding, trained, encoded, and decoded step by step.

• Fullstack AI Engineering for Web Developers

  Level up your web apps by building custom RAG chatbots, autonomous workflow agents, and your own trained neural networks from scratch. Let's turn your code into intelligent systems. * **Build custom chatbots** powered by your own data via RAG * **Develop autonomous AI agents** to automate complex user workflows * **Integrate smart AI APIs** directly into fullstack web applications * **Train and deploy neural networks** from scratch with ease

• Building Product Assistants with ADK

  Turn your web apps into intelligent products by mastering memory, doc search, and API-calling tools to build and deploy reliable, production-ready AI assistants with confidence. * **Ground responses** using your product docs and data * **Connect external APIs** for automated agent tool use * **Manage persistent memory** and multi-turn user sessions * **Evaluate and deploy** production-ready agents with confidence

• AI Math Tutoring Evidence

  Bridge learning science with cutting-edge LLM research to evaluate what actually works in K-12 instruction, empowering you to shape high-impact, evidence-based math tutoring products. * **Empirical evaluation** of recent K-12 math AI research * **Strategic roadmapping** for evidence-based EdTech product decisions * **Limitation analysis** targeting LLM gaps in mathematical reasoning * **Learning science integration** for effective generative tutoring interventions

• iPAS AI應用規劃師初級衝刺

  憑藉你的 AWS 基礎，這堂課將精準鎖定兩大考科核心與精選試題，助你在 11 月 7 日高效通關取證。 * **掌握雙科核心概念**：對齊兩大考綱關鍵要點 * **無縫接軌證照考點**：善用雲端基礎轉化本土題型 * **精準破解歷屆試題**：實戰演練掌握出題邏輯 * **穩拿商業應用分數**：結合管理思維落實技術規劃

• How Machine Learning Models Learn

  For curious beginners and early-career engineers: how models learn from data, from features, loss, and gradient descent to overfitting, honest evaluation, and the model families that lead to neural networks and today's language models.

• Building LLM Applications: Architecture to Production

  For software engineers shipping LLM features in real products: how to pick between workflows and agents, get typed outputs and reliable tool calls, build evals you can trust, control cost and latency, and handle failures and prompt injection in production.

• AI Product Management: Scope, Evals, Launch

  For product managers moving into AI products. You will learn when AI beats a simple rule, how to spec error costs and success metrics, build evals without coding, weigh model cost against quality, and launch with real safeguards.

• No-Code AI Automation for Business

  For non-technical professionals and small business owners who want to hand repetitive work to tools like Zapier, Make, and Power Automate, and add AI steps, human review, error handling, and data safeguards that keep those automations reliable and worth the time.

• 從零建置 Orca AI 個人助理總控（完整實作版）

  從零理解 CLI 代理、git worktree 與 Orca，照每章的實作步驟，在自己的電腦搭出一套 AI 個人助理總控：用 worktree 隔離多個專案，由總控判斷派工、驗收與覆核，用任務帳本和收據追蹤每件事，設好安全邊界，最後把課業、信件、行事曆、提醒等生活與工作的大小事接進同一套流程。學員背景：用過 ChatGPT 或 Claude 聊天，會基本終端機和 git，沒用過 CLI 代理、worktree 或 Orca。

• 多代理工作流系統化設計

  專為進階工程師設計，帶你系統化拆解專家經驗為審查準則與對抗式流程，打造高品質、免盯盤的高效多代理協作架構。 * **模組化專家拆題**：精準解構複雜邏輯，產出高覆用派工範本 * **對抗式審查升級**：建立多攻擊者與裁判機制，阻絕產出返工 * **專家直覺準則化**：將工程判斷轉為機器可讀的嚴格檢查清單 * **全自動放手標準**：設計信心度閘門與自動化交卷流程，徹底省下盯盤時間

• AI Data Annotation Zero-to-One: Track Selection, RLHF Execution, and QA Management

  帶你拆解主流標註賽道並實作RLHF，從零建立技能定位，掌握品質管理晉升專案負責人。 * **賽道適配矩陣分析**：鎖定最佳個人優勢賽道 * **三大標註實戰操作**：精通 CV、NLP 與 RLHF 規範 * **品質檢驗標準掌握**：熟練 3H 原則與邊界案例處理 * **專案管理躍升能力**：制定 SOP 與 IAA 質檢校準

• Fine-Tuning Small Language Models

  Welcome! You will master hands-on techniques to adapt compact AI models for custom tasks, giving you powerful practical skills for your upcoming academic and personal projects. * **Preparing custom datasets** for optimal training * **Applying parameter-efficient techniques** like LoRA and QLoRA * **Training compact models** on consumer hardware * **Evaluating performance** of custom lightweight architectures

• Photonics for Neural Network Inference

  Explore how light-speed matrix operations can accelerate AI inference, analyze the hardware bottlenecks holding photonic chips back, and master the real-world engineering trade-offs beyond silicon. * **Optical neural network fundamentals**: core light-based computing concepts * **Hardware-accelerated inference**: matrix multiplication using optical physics * **Nonlinear operational limits**: photonics-friendly functions versus scaling bottlenecks * **Engineering and scaling tradeoffs**: critical adoption and manufacturing hurdles